
Detail-oriented MSc candidate in Artificial Intelligence with experience developing production-level machine learning systems, specialising in computer vision, deep learning and time-series modelling. Skilled in Python and contemporary ML frameworks, delivering measurable performance improvements and end-to-end data pipelines. Seeking a Machine Learning Engineer position to implement reliable, data-driven solutions.
YOLO Object Detection & Synthetic Dataset Generation
• Built an object detection model with YOLO and PyTorch that reached around 90% accuracy on a custom dataset.
• Generated 4,000+ synthetic training images with automated annotation to overcome limited real-world data, then improved results through data augmentation and model tuning.
Blockchain-Enabled Hyperlocal Air Quality Monitoring
• Designed a privacy-preserving IoT system for street-level air monitoring in Bristol, using LoRaWAN sensors, Raspberry Pi edge gateways and a random forest model to calibrate low-cost sensors locally so raw data never leaves the neighbourhood.
• Added a blockchain audit layer for tamper-proof data-access logging and a generative design (genetic algorithm) solver to optimise sensor placement for health coverage and privacy, at ~1/20th the cost of a reference station.
Malicious URL Detector (Django)
• Built a web app classifying URLs as safe or malicious, integrating the VirusTotal API for real-time threat analysis.